Box 1. Honing in on differentiation.

A McMaster University team led by Mick Bhatia has used a human pluripotent stem cell (hPSC) line that reproduces neoplastic properties of somatic cancer stem cells14 to develop a screening platform to identify compounds able to selectively target neoplastic cells and not normal hPSCs.15

Cancer stem cells are characterized by enhanced self-renewal with limited cellular differentiation, which drives tumor growth. However, the lack of a robust in vitro assay to interrogate human cancer stem cell differentiation has made it difficult to develop therapeutics that interrupt the process.

Bhatia's team set out to develop such an assay. He is professor of biochemistry and biomedical sciences at McMaster University and director and senior scientist at the McMaster Stem Cell and Cancer Research Institute.

The group used neoplastic hPSCs and normal hPSCs expressing GFP reporters to screen compound libraries. A decrease in the reporters indicated a loss of self-renewal and an induction of differentiation.

Hits from the screen included thioridazine, fluphenazine, prochlorperazine, rapamycin and Teva Pharmaceutical Industries Ltd.'s lestaurtinib.

Thioridazine, fluphenazine and prochlorperazine are dopamine receptor antagonists. Rapamycin is a mammalian target of rapamycin (mTOR; FRAP; RAFT1) inhibitor, and the tyrosine kinase inhibitor lestaurtinib is in Phase I/II testing for myeloproliferative disorder.

The team next set out to validate that thioridazine could block cancer progression in vivo. In mice, thioridazine decreased leukemic stem cell engraftment compared with vehicle.

"Using a 2,446-compound library, our screening platform provided us with 26 possible hits, with thioridazine being attractive to us because it exhibited the lowest EC50 in neoplastic stem cells without affecting normal stem cells and is a known antipsychotic with FDA approval," said Bhatia. "Although our library didn't contain trifluoperazine itself-Narla's hit-our screen did identify a couple of trifluoperazine analogs as potential anticancer therapeutics."

He added, "I think our screening platform is very amenable for identifying anticancer therapeutics. Typical assays show if compounds affect cancerous stem cells" but do not show what happens to normal cells. "Our platform can inform you if a compound will affect only neoplastic cells and spare normal cells."

McMaster University has mature provisional patents for the screening technology, which is licensed to Actium Research Inc.